Semantics- oriented systems
They include the logical-analytical systems, which provide the automatic extraction of semantic information from documents and their use for the solution of the problems, which relate to the class of intellectual.
Semantics-oriented systems are the new class of systems. Similar systems for different fields of application were developed within the framework of commercial projects of IPI RAN (Institute for Informatics Problems Russian Academy of Science): DIEZ, Criminal, LINGVO- MASTER and other
For this aims the tools of construction Knowledge Base (KB) and corresponding procedures of the meaningful texts processing were created. We transfer of complex linguistic analysis to the level of processing knowledge. Obtained results are corresponded the newest scientific research abroad.
The systems are oriented to process the texts of natural language with the delivery to the user (analyst) the necessary information in the most understandable form. Standard tasks of the systems:
- Automatic introduction of documents with their division on the part and lexical analysis;
- Automatic formalization of text information (texts in Russian and English) with the creation of s own KB;
- Search for similar documents and mentioned in them (information) objects on the basis of the criterion of their semantic proximity;
- Answer to the questions in the natural language;
- Search of objects according to the connections with other objects;
- Using of statistical methods for the search for dependences and their representation in the form of temporary diagrams;
- Automatic forming of information in Data Bases.
Systems relate to the class of logical-analytical. These are the complex programs, which automate the process of introduction, formalization and the analysis of text documents, their use in the tasks of search and subsequent analysis. Formalization consists in extraction of the significant information: the interesting user objects (for example, FIO of the persons, name organizations, cities and others), and also their signs, connections, the attributes (this they can be property, verbal portraits and other). They form the structures of knowledge in KB. Each system contains its own KB, which provide:
- Storage of the significant information and connections;
- Effective search and data analysis according to the connections.
Thus, we use the principles and tools of the construction of the new class of the systems, based on the knowledge structures.